* Update backend
* Update frontend
* Keep old plugin id
* Update docs
* Place doc images to a new directory
* Legacy support for stackdriver-auto alignment
* Consistent plugin name
* Apply suggestions from code review
Co-authored-by: Marcus Efraimsson <marcus.efraimsson@gmail.com>
Co-authored-by: Tobias Skarhed <1438972+tskarhed@users.noreply.github.com>
* Update docs
* Update public/app/plugins/datasource/cloud-monitoring/README.md
Co-authored-by: Marcus Efraimsson <marcus.efraimsson@gmail.com>
* Add reference to the data source formerly being named Stackdriver
* Update pkg/models/datasource.go
Co-authored-by: Carl Bergquist <carl@grafana.com>
* Fix gofmt
Co-authored-by: Marcus Efraimsson <marcus.efraimsson@gmail.com>
Co-authored-by: Tobias Skarhed <1438972+tskarhed@users.noreply.github.com>
Co-authored-by: Carl Bergquist <carl@grafana.com>
Profiling and tracing can now be enabled/disabled separately.
Adds argument for trace file path.
Support overriding profiling and tracing argument using
environment variables.
Update docs.
Fixes#20576
* pkg/cmd: Check errors
* pkg/cmd: Make sure server waits on services, even in case of error
* pkg/cmd: Inform of error to show help
* pkg/cmd: Only warn on failure to send systemd notification
* pkg/cmd: Don't log errors stemming from context cancelation
* pkg/cmd: Don't fail if unable to write to systemd
* Rename GrafanaServerImpl to Server
* Make flag dependencies explicit
* Extract method for building service graph
* Document and clarify methods
* Rename HttpServer to HTTPServer
See,
$ gometalinter --vendor --deadline 10m --disable-all --enable=errcheck ./...
pkg/cmd/grafana-server/server.go:77:22⚠️ error return value not checked (serviceGraph.Provide(&inject.Object{Value: bus.GetBus()})) (errcheck)
pkg/cmd/grafana-server/server.go:78:22⚠️ error return value not checked (serviceGraph.Provide(&inject.Object{Value: g.cfg})) (errcheck)
pkg/cmd/grafana-server/server.go:79:22⚠️ error return value not checked (serviceGraph.Provide(&inject.Object{Value: routing.NewRouteRegister(middleware.RequestMetrics, middleware.RequestTracing)})) (errcheck)
pkg/cmd/grafana-server/server.go:80:22⚠️ error return value not checked (serviceGraph.Provide(&inject.Object{Value: cache.New(5*time.Minute, 10*time.Minute)})) (errcheck)
pkg/cmd/grafana-server/server.go:87:23⚠️ error return value not checked (serviceGraph.Provide(&inject.Object{Value: service.Instance})) (errcheck)
pkg/cmd/grafana-server/server.go:90:22⚠️ error return value not checked (serviceGraph.Provide(&inject.Object{Value: g})) (errcheck)
* Add new build info metrics that contains more info
The goal was to add more information about Grafana. But rather than
just adding those to the current metrics I created a new metric
since its a common pattern in the prometheus community to expose that
info in a metric named `*_build_info`.
We keep the old metric to avoid introducing any breaking changes but
we should be able to remove it the next breaking change
See,
$ gometalinter --vendor --deadline 10m --disable-all --enable=megacheck ./...
pkg/cmd/grafana-server/main.go:103:42⚠️ os.Kill cannot be trapped (did you mean syscall.SIGTERM?) (SA1016) (megacheck)
* Reopen log files after receiving a SIGHUP signal
Implements a system signals listener in log.FileLogWriter.
After receiving a SIGHUP signal from the system, the listener will close the current log file and then open it again.
The listener will finishes after receiving a SIGINT, SIGKILL or SIGTERM signal.
Closesgrafana/grafana#2497
* Move the SIGHUP handling to the main file
